Key points are not available for this paper at this time.
Abstract It is widely assumed that long-wavelength infrared (IR) emission from cold dust ( T ∼ 20–40 K) is a reliable tracer of star formation even in the presence of a bright active galactic nucleus (AGN). Based on radiative transfer (RT) models of clumpy AGN tori, hot dust emission from the torus contributes negligibly to the galaxy spectral energy distribution (SED) at λ ≳ 100 μ m. However, these models do not include AGN heating of host-galaxy-scale diffuse dust, which may have far-IR (FIR) colors comparable to cold diffuse dust heated by stars. To quantify the contribution of AGN heating to host-galaxy-scale cold dust emission at λ ≳ 100 μ m, we perform dust RT calculations on a simulated galaxy merger both including and excluding the bright AGN that it hosts. By differencing the SEDs yielded by RT calculations with and without AGNs that are otherwise identical, we quantify the FIR cold dust emission arising solely from reprocessed AGN photons. In extreme cases, AGN-heated host-galaxy-scale dust can increase galaxy-integrated FIR flux densities by factors of 2–4; star formation rates calculated from the FIR luminosity assuming no AGN contribution can overestimate the true value by comparable factors. Because the FIR colors of such systems are similar to those of purely star-forming galaxies and redder than torus models, broadband SED decomposition may be insufficient for disentangling the contributions of stars and heavily dust-enshrouded AGNs in the most IR-luminous galaxies. We demonstrate how kiloparsec-scale resolved observations can be used to identify deeply dust-enshrouded AGNs with cool FIR colors when spectroscopic and/or X-ray detection methods are unavailable.
McKinney et al. (Fri,) studied this question.